- Email: [email protected]
Corneal volume, pachymetry, and correlation of anterior and posterior corneal shape in subclinical and different stages of clinical keratoconus
ARTICLE
Corneal volume, pachymetry, and correlation of anterior and posterior corneal shape in subclinical and different stages of clinical keratoconus David P. Pin˜ero, MSc, Jorge L. Alio´, MD, PhD, Alicia Aleso´n, OD, Munir Escaf Vergara, MD, Mauricio Miranda, MD
PURPOSE: To evaluate the corneal volume, pachymetry, and correlation of anterior and posterior corneal shape in subclinical and clinical keratoconus. SETTING: Vissum Corporation, Alicante, Spain. METHODS: Eyes were placed into 1 of 4 groups as follows: keratoconus 2 (grade II), keratoconus 1 (grade I), subclinical (subclinical keratoconus), and control (normal eyes). All eyes had an ophthalmologic examination including corneal evaluation (curvature, elevation, asphericity, pachymetry, corneal volume) by rotating Scheimpflug imaging (Pentacam). The posterior–anterior corneal power ratio was also calculated. RESULTS: Seventy-one eyes (51 patients; aged 16 to 64 years) were evaluated. Astigmatism and keratometry of both corneal surfaces were statistically significantly higher in the keratoconus 1 and 2 groups (P%.02). Posterior astigmatism was statistically significantly higher in the subclinical group than in the control group (P Z .01). A strong correlation (rR 0.81) was found between anterior and posterior curvature in the normal and subclinical groups; the correlation was weaker in clinical keratoconus cases (r% 0.56). The correlation in astigmatism between the anterior and posterior surface was good in all keratoconus groups (rR0.81). The posterior–anterior corneal power ratio was significantly higher in the keratoconus 2 group than in the other groups (P%.01). Pachymetric readings were progressively lower in eyes with subclinical, early, or moderate keratoconus (P<.01). The corneal volume was statistically significantly lower in the keratoconus 2 group than in the other groups (P Z .04). CONCLUSION: The correlation between anterior and posterior corneal curvature was lower in keratoconus, although the correlation between anterior and posterior astigmatism was maintained. Financial Disclosure: No author has a financial or proprietary interest in any material or method mentioned. J Cataract Refract Surg 2010; 36:814–825 Q 2010 ASCRS and ESCRS
Keratoconus is an ectatic corneal disorder characterized in most cases by progressive corneal thinning that results in corneal protrusion, irregular astigmatism, and decreased vision.1 Its incidence varies depending on several factors, such as ethnicity and the criteria used to establish the diagnosis; most estimates place the incidence in the general population between 50 and 230 per 100 000.1 Corneal topography is a valuable tool for confirming the keratoconus diagnosis.1 Significant corneal steepening in the anterior corneal surface of keratoconic eyes is always observed; the steepening is usually confined to 1 or 2 quadrants.1,2 Therefore, detecting moderate and advanced 814
Q 2010 ASCRS and ESCRS Published by Elsevier Inc.
keratoconus is not difficult using corneal topography and biomicroscopic, retinoscopic, and pachymetric findings.1 Detection is difficult with very early or preclinical stages of this ectatic disorder. The terms forme fruste keratoconus, subclinical keratoconus, and keratoconus suspect have been used to designate early stages of keratoconus that do not show on biomicroscopy but have subtle topographic features similar to those of clinical keratoconus.3–8 The biomicroscopic findings in keratoconus, such as stromal thinning and posterior stress lines, suggest that the posterior corneal shape is altered, perhaps independently of the anterior surface. Tomidokoro et al.9 0886-3350/10/$dsee front matter doi:10.1016/j.jcrs.2009.11.012
CLINICAL AND SUBCLINICAL KERATOCONUS
found that both anterior curvature and posterior curvature were affected in eyes with keratoconus as well as in keratoconus-suspect eyes. Significantly large best-fit sphere (BFS) values3,5,10 and high posterior elevations4,5,11 have been observed in eyes with clinical or subclinical keratoconus. Thus, manifestations of keratoconus occur at the posterior corneal surface, even in early stages of the disease, and identifying these changes could help clinicians detect subclinical keratoconus. To date, studies of posterior changes in eyes with keratoconus were performed using scanning-slit technology, which has been shown to be inaccurate for some posterior corneal measurements in eyes that have had laser in situ keratomileusis (LASIK).12–16 A fairly recent advance in corneal topography was the introduction of Scheimpflug photography for corneal topographic characterization; this technique allows the study of both the anterior and posterior corneal surfaces17 and provides more repeatable and reproducible anterior and posterior measurements of corneal power than scanning-slit technology. Kawamorita et al.18 reported 0.19 of agreement for withinrater consecutive measurements of posterior corneal power measurements of diopter (D) with a Scheimpflug-based system and 0.96 D with a scanning-slit system and for between-rater measurements, of 0.56 D and 1.58 D, respectively. The aim of the present study was to evaluate changes in the anterior and posterior corneal surfaces, pachymetry, and corneal volume in eyes with subclinical keratoconus or grade I or II clinical keratoconus using a Scheimpflug imaging system. We also analyzed the degree of correlation between the 2 corneal surfaces from the early stages of the disorder and
Submitted: March 4, 2009. Final revision submitted: November 20, 2009. Accepted: November 23, 2009.
815
the contribution of the correlation to volumetric modifications. PATIENTS AND METHODS This comparative study enrolled patients examined at Vissum Instituto Oftalmolo´gico de Alicante, Spain, or Instituto Oftalmosalud, Peru, between January 2008 and July 2008. All patients were informed about inclusion in the study and provided informed consent in accordance with the Declaration of Helsinki. An ethics board committee approved the study.
Patient Grouping Based on an examination, eyes were placed into 1 of 4 groups as follows: keratoconus 2 (grade II), keratoconus 1 (grade I), subclinical (subclinical keratoconus), and control (normal eyes). Clinical keratoconus was defined as evident findings characteristic of keratoconus (eg, corneal topography with asymmetric bow-tie pattern with or without skewed axes) (Figures 1 and 2), and at least 1 keratoconus sign (eg, stromal thinning, conical protrusion of the cornea at the apex, Fleischer ring, Vogt striae, or anterior stromal scar) on slitlamp examination.1 The Alio´ and Shabayek classification19 was used to grade the keratoconus. In the classification, grade II keratoconus (Figure 1) is characterized by a mean keratometry (K) value less than 53.00 D and a coma-like (ie, 3rd-, 5th-, and 7th-order Zernike terms) root-mean-square (RMS) value between 2.5 mm and 3.5 mm. Grade I keratoconus (Figure 2) is characterized by a mean K value less than 48.00 D and a coma-like RMS value between 1.50 mm and 2.50 mm. Subclinical keratoconus was diagnosed using criteria defined in previous studies,3–10,18,20–22 including corneal topography with abnormal localized steepening or an asymmetric bow-tie pattern (Figure 3), a normal-appearing cornea on slitlamp biomicroscopy, and at least 1 of the following signs: steep keratometric curvature (O47.00 D), oblique cylinder greater than 1.50 D, central corneal thickness less than 500 mm, or clinical keratoconus in the fellow eye. Eyes were considered normal if they had no ocular pathology, no previous ocular surgery, no significant refractive error, and no irregular corneal pattern (Figure 4). In this group, only 1 eye of each patient was evaluated (random sampling).
Patient Assessment
From the Keratoconus Unit (Pin˜ero, Alio´, Aleso´n), Vissum Corporation–Instituto Oftalmologico de Alicante, Departamento de O´ptica (Pin˜ero), Farmacologı´a y Anatomı´a, Universidad de Alicante, Division of Ophthalmology (Alio´), Universidad Miguel Herna´ndez, Alicante, Spain; Instituto Oftalmosalud SRL (Vergara, Miranda), Lima, Peru´. Supported in part by a grant from the Spanish Ministry of Health, Instituto Carlos III, Red Tema´tica de Investigacio´n Cooperativa en Salud Patologı´a Ocular del Envejecimiento, Calidad Visual y Calidad de Vida, Subproyecto de Calidad Visual (RD07/0062). Corresponding author: David P. Pin˜ero, MSc, Vissum Corporation– Instituto Oftalmologico de Alicante, Avenida de Denia s/n, Edificio Vissum, 03016 Alicante, Spain. E-mail: [email protected].
All patients had a complete ophthalmologic examination that included uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), manifest refraction, and slitlamp and fundus evaluations. Corneal evaluation was performed using a Scheimpflug imaging system (Pentacam, Oculus Optikgera¨te GmbH). This noninvasive system measures and characterizes the anterior segment.23 A rotating Scheimpflug camera captures 100 images with 500 measurement points on the anterior and posterior corneal surface over a 180-degree rotation. The elevation data from all images are combined to form a 3-dimensional (3-D) reconstruction of the corneal structure. After processing the information, the internal software provides several calculations and parameters. In the current study, software version 6.02r10 was used with a previously described measurement protocol.24 Several studies23–26 report that the system’s
J CATARACT REFRACT SURG - VOL 36, MAY 2010
816
CLINICAL AND SUBCLINICAL KERATOCONUS
Figure 1. Corneal topographic analysis of an eye with keratoconus grade II (Scheimpflug system). Left: Anterior axial topographic map. Right: Posterior axial topographic maps. The corneal geometry of the anterior corneal surface and posterior corneal surface is similar, with the same topographic pattern.
pachymetric and geometric measurements (curvature, asphericity, elevation) have good repeatability. The following anterior and posterior corneal surface parameters were evaluated with the Scheimpflug system: corneal dioptric power in the flattest meridian in the 3.0 mm central zone, corneal dioptric power in the steepest meridian in the 3.0 mm central zone, mean corneal power in the 3.0 mm zone, corneal astigmatism in the 3.0 mm zone, the BFS in an 8.0 mm diameter analysis area, and mean asphericity in an 8.0 mm diameter corneal area. Central pachymetry, minimum pachymetry, and corneal volume were also recorded and analyzed. In addition, the posterior–anterior corneal power ratio and posterior– anterior astigmatism ratio were calculated; that is, the ratios of the anterior value to the posterior value of the corresponding variable (ie, corneal power or astigmatism).
Statistical Analysis The number of patients in the study was chosen based on sample-size calculations obtained using a standard procedure. Considering means and standard deviations in previous studies3,5,9,17,23,24 of keratoconus diagnosis and assuming a specific significance level and statistical power (80%), the sample-size was estimated using Ene 2.0 software (GlaxoSmithKline). Statistical analysis was performed using SPSS software for Windows (version 15.0, SPSS, Inc.). Normality of all data
samples was first checked by the Shapiro-Wilks test. When parametric analysis was possible, 1-way analysis of variance (ANOVA) with Bonferroni post hoc comparison was used to compare visual, refractive, topographic, pachymetric, and volumetric parameters between the 4 groups of eyes. If variances were not homogeneous (checked by Levene test), Tamhane post hoc analysis was used. In all cases, differences were considered statistically significant when the P value was less than 0.05. When parametric analysis was not possible, the Kruskal-Wallis test was used for between-group comparisons and a P value less than 0.05 was considered statistically significant. Post hoc analysis was by the MannWhitney test with Bonferroni adjustment. Correlation coefficients (Pearson or Spearman depending on whether normality could be assumed) were used to assess the correlation between mean keratometry and astigmatism of the anterior corneal surface and posterior corneal surface in each of the 4 groups of eyes. In addition, linear regression analysis was performed to determine the relationship between anterior parameters and posterior parameters and to determine significant correlations between parameters. An analysis of the receiver operating characteristic (ROC) curve of Scheimpflug system parameters was performed to obtain cutoff values and thus detect keratoconus and keratoconus-suspect cases. This analysis was performed only for corneal parameters for which there was a statistically significant difference between control eyes and eyes with keratoconus or between control eyes and eyes with
Figure 2. Corneal topographic analysis of an eye with keratoconus grade I (Scheimpflug system). Left: Anterior axial topographic map. Right: Posterior axial topographic maps. The corneal geometry of the anterior corneal surface and posterior corneal surface is similar, with the same topographic pattern.
J CATARACT REFRACT SURG - VOL 36, MAY 2010
817
CLINICAL AND SUBCLINICAL KERATOCONUS
Figure 3. Corneal topographic analysis of an eye with subclinical keratoconus (Scheimpflug system). Left: Anterior axial topographic map. Right: Posterior axial topographic maps. The corneal geometry of the anterior corneal surface and posterior corneal surface is similar, with the same topographic pattern.
subclinical keratoconus. Cutoff values for detection of keratoconus (discrimination between eyes with keratoconus and control eyes) were calculated for all Scheimpflug system parameters except the ratios. Cutoff values to detect subclinical keratoconus cases (discrimination between eyes with subclinical keratoconus and normal eyes) were calculated for anterior and posterior corneal astigmatism in the 3.0 mm zone. These cutoff values corresponded to the points of the curve with the highest accuracy (ie, least false negative and false positive results). In addition, the area under the ROC curve was calculated to measure test accuracy, with an area of 1.0 representing a perfect test and an area of 0.5 representing an unusable test. The closer the curve follows the left-hand border and then the top border of the ROC space, the more accurate the test; that is, the test can identify more true positives while minimizing the number of false positives.
RESULTS The study evaluated 71 eyes of 51 patients ranging in age from 16 to 64 years. There were 37 right eyes (52.11%), 34 left eyes (47.89%), 29 men (56.86%), and 22 women (43.14%). No patient had a history of contact lens wear, eye disease, or surgery. The keratoconus 2 group comprised 18 eyes with a mean age of 39.06 years G 11.66 (SD). The
keratoconus 1 group comprised 19 eyes with a mean age of 33.58 G 11.26 years. The subclinical group comprised 14 eyes with a mean age of 30.00 G 9.15 years; all subclinical cases were unilateral. The control group comprised 20 eyes with a mean age of 32.30 G 6.59 years. No cone opacity was observed in any eye. Visual Acuity and Refraction Table 1 shows the visual acuity and refraction by group. Statistically significant differences in sphere and cylinder were found between groups (P%.01, Kruskal-Wallis test). Specifically, cylinder was statistically significantly higher in the keratoconus 1 and keratoconus 2 groups and sphere in the keratoconus 2 group (all P!.01, Mann-Whitney test). The only statistically significant difference between the control group and the subclinical group was in cylinder (P Z .03, Mann-Whitney test). There were no statistically significant differences in sphere or cylinder between the keratoconus 1 group and keratoconus 2 group (all PR.18, Mann-Whitney test). The difference in CDVA between the 4 groups was statistically significant (P!.01, Kruskal-Wallis test).
Figure 4. Corneal topographic analysis of a normal eye (Scheimpflug system). Left: Anterior axial topographic map. Right: Posterior axial topographic maps. There are differences between the anterior and posterior topographic patterns, especially in the magnitude of astigmatism.
J CATARACT REFRACT SURG - VOL 36, MAY 2010
818
CLINICAL AND SUBCLINICAL KERATOCONUS
Table 1. Visual acuity and refractive outcomes. Group Parameter Sphere (D) Mean G SD Range Cylinder (D) Mean G SD Range SE (D) Mean G SD Range CDVA Mean G SD Range
Control
Subclinical
Keratoconus 1
Keratoconus 2
1.26 G 2.54 6.00 to C4.00
1.16 G 1.15 3.25 to 0.00
3.13 G 4.19 C1.75 to 13.75
5.36 G 4.56 17.00 to C2.00
0.54 G 1.00 3.50 to 0.00
1.21 G 1.05 3.25 to 0.00
3.80 G 2.16 8.50 to 1.00
4.28 G 2.20 9.00 to 0.00
1.77 G 1.35 4.88 to 0.00
5.03 G 4.13 14.75 to C0.63
7.50 G 4.14 18.00 to 0.25
1.02 G 0.13 0.80 to 1.20
0.80 G 0.19 0.33 to 1.00
0.44 G 0.20 0.13 to 0.80
P Value .01
!.01
!.01 1.48 G 2.43 7.00 to C2.25
!.01 1.12 G 0.10 1.00 to 1.20
CDVA Z corrected distance visual acuity; SE Z spherical equivalent
Curvature Table 2 shows the curvature and elevation data for the anterior corneal surface by group. There were statistically significant between-group differences in the flattest and steepest meridians, mean corneal power, and corneal astigmatism in the 3.0 mm central zone; the BFS; and the mean asphericity in an 8.0 mm diameter area (P!.01, 1-way ANOVA and Kruskal-Wallis tests). Specifically, the corneal dioptric power in the steepest meridian and mean corneal power in the 3.0 mm central zone were statistically significantly higher in the keratoconus 2 and keratoconus 1 groups than in the subclinical and control groups (P!.01, Mann-Whitney test). The corneal dioptric power in the flattest meridian in the 3.0 mm central zone was significantly higher in the keratoconus 2 and keratoconus 1 groups than in the control and subclinical groups and in the keratoconus 1 group than in the control group (P!.01 and P Z .04, respectively; Bonferroni test). The anterior corneal astigmatism was statistically significantly higher in the keratoconus 2, keratoconus 1, and subclinical groups than in the control group (P!.01, Mann-Whitney test). The BFS value was statistically significant lower in the keratoconus 2 group (P!.01, Bonferroni test); the difference between the subclinical group and keratoconus 1 group was not significant (P Z .99, Bonferroni test). The anterior corneal surface was significantly more prolate in the keratoconus 2 group and keratoconus 1 group than in the other 2 groups (P!.01, Bonferroni test). Table 3 shows the curvature and elevation data for the posterior corneal surface by group. The corneal dioptric power in the flattest and steepest meridians, mean corneal power, and corneal astigmatism in the 3.0 mm central zone were statistically significantly
higher in the keratoconus 2 and keratoconus 1 groups than in the subclinical and control groups (P%.02, Bonferroni and Mann-Whitney tests). Statistically significant differences in posterior astigmatism between the control group and subclinical group were also found (P Z .01, Mann-Whitney test). The posterior BFS value was statistically significantly lower in the keratoconus 2 group than in the other 3 groups (P%.02, Mann-Whitney test); there were no significant differences in the BFS between the subclinical group and the keratoconus 1 group (P Z .24, MannWhitney test). The posterior corneal surface was significantly more prolate (marked negative asphericity) in the keratoconus 2 and keratoconus 1 groups than in the other 2 groups (P!.01, Mann-Whitney test). Relationship Between Anterior and Posterior Corneal Surfaces Table 4 shows the correlation coefficients between the curvatures, BFS, and asphericities of anterior and posterior corneal surfaces by group. In the control group, a significant and strong correlation was found between the anterior curvature and posterior curvature (Figure 5). However, the correlation was lower in the other 3 groups, especially in the keratoconus 1 group, in which the correlation between the anterior and posterior mean keratometric values was poor. Regarding astigmatism, there was a good correlation between the anterior surface and posterior surface in all groups except the control group (Figure 6). There was a strong correlation between the anterior BFS and posterior BFS in the control group and subclinical group. Regarding corneal asphericity, the correlation between the anterior surface and posterior surface was good in the keratoconus 1 group, moderate in
J CATARACT REFRACT SURG - VOL 36, MAY 2010
819
CLINICAL AND SUBCLINICAL KERATOCONUS
Table 2. Curvature, asphericity, and elevation data for the anterior corneal surface by group. Group Parameter Corneal power* (D) Flattest meridian Mean G SD Range Steepest meridian Mean G SD Range Mean corneal power* (D) Mean G SD Range Corneal astigmatism* (D) Mean G SD Range BFS† (mm) Mean G SD Range Mean asphericity† Mean G SD Range
Control
Subclinical
Keratoconus 1
Keratoconus 2
42.82 G 1.69 40.10 to 46.30
43.00 G 1.31 40.90 to 45.40
44.27 G 1.30 42.50 to 46.30
48.27 G 2.06 44.90 to 52.50
43.81 G 1.97 40.80 to 47.00
44.92 G 1.77 41.60 to 47.60
49.05 G 1.07 47.80 to 52.00
53.09 G 2.20 48.60 to 56.30
43.29 G 1.74 40.40 to 46.60
43.93 G 1.47 41.20 to 46.50
46.51 G 0.70 45.30 to 47.60
50.52 G 1.60 48.10 to 53.60
1.13 G 0.84 0.40 to 3.70
1.93 G 1.02 0.40 to 3.80
4.55 G 2.11 0.90 to 9.30
4.83 G 2.83 0.80 to 10.30
7.89 G 0.31 7.32 to 8.43
7.74 G 0.30 7.08 to 8.18
7.64 G 0.23 7.27 to 8.04
7.26 G 0.30 6.70 to 7.75
0.29 G 0.09 0.46 to 0.14
0.33 G 0.23 0.67 to 0.01
0.65 G 0.27 1.33 to 0.26
1.18 G 0.32 1.68 to 0.58
P Value
!.01
!.01
!.01
!.01
!.01
!.01
BFS Z best-fit sphere *Central 3.0 mm zone † 8.0 mm diameter area
Table 3. Curvature, asphericity and elevation data for the posterior corneal surface by group. Group Parameter Corneal power* (D) Flattest meridian Mean G SD Range Steepest meridian Mean G SD Range Mean corneal power* (D) Mean G SD Range Corneal astigmatism* (D) Mean G SD Range BFS† (mm) Mean G SD Range Mean asphericity† Mean G SD Range
Control
Subclinical
Keratoconus 1
Keratoconus 2
P Value
!.01 6.15 G 0.31 6.70 to 5.60
6.16 G 0.30 6.90 to 5.80
6.53 G 0.41 7.70 to 6.00
7.22 G 0.61 8.30 to 6.00
6.48 G 0.42 7.20 to 5.90
6.73 G 0.44 7.40 to 6.10
7.27 G 0.62 8.30 to 5.60
8.29 G 0.43 9.10 to 7.40
6.31 G 0.33 6.90 to 5.80
6.42 G 0.33 7.10 to 6.00
6.86 G 0.23 7.40 to 6.40
7.69 G 0.46 8.70 to 7.00
0.33 G 0.19 0.10 to 0.70
0.56 G 0.28 0.30 to 1.20
1.07 G 0.50 0.30 to 1.90
1.07 G 0.60 0.30 to 2.50
6.48 G 0.31 5.89 to 7.04
6.34 G 0.28 5.64 to 6.65
6.25 G 0.31 5.91 to 7.26
6.14 G 0.76 5.42 to 8.54
0.66 G 0.33 1.40 to 0.04
1.17 G 0.42 1.81 to 0.14
!.01
!.01
!.01
.08
!.01 0.34 G 0.24 1.01 to 0.02
0.36 G 0.33 1.03 to 0.05
BFS Z best-fit sphere *Central 3.0 mm zone † 8.0 mm diameter area
J CATARACT REFRACT SURG - VOL 36, MAY 2010
820
CLINICAL AND SUBCLINICAL KERATOCONUS
Table 4. Correlation coefficients for curvature, asphericity and elevation parameters between the anterior corneal surface and posterior corneal surface by group. Group Parameter Corneal power* (D) Flattest meridian r value P value Steepest meridian r value P value Mean corneal power* (D) r value P value Corneal astigmatism* (D) r value P value BFS† (mm) r value P value Mean asphericity† r value P value
Control
Subclinical
Keratoconus 1
Keratoconus 2
0.90 !.01
0.83 !.01
0.37 .12
0.73 .01
0.96 !.01
0.81 !.01
0.32 .19
0.67 !.01
0.94 !.01
0.85 !.01
0.34 .16
0.56 .02
0.62 !.01
0.85 !.01
0.81 .01
0.91 !.01
0.92 !.01
0.96 !.01
0.65 !.01
0.56 .02
0.17 .47
0.62 .02
0.89 !.01
0.63 !.01
BFS Z best-fit sphere *Central 3.0 mm zone † 8.0 mm diameter area
the subclinical and keratoconus 2 groups, and very poor in the control group (Table 4). The mean posterior–anterior corneal power ratio was 0.146 G 0.003 in the control group, 0.146 G 0.005 in the subclinical group, 0.148 G 0.005 in the keratoconus 1 group, and 0.152 G 0.007 in the keratoconus 2 group. The ratio was statistically significantly higher in the keratoconus 2 group than in the other 3 groups. The mean posterior–anterior astigmatism ratio was 0.333 G 0.170, 0.338 G 0.202, 0.268 G 0.188, and 0.257 G 0.134, respectively. There were no statistically significant differences in the ratio between the groups (P Z .12, Kruskal-Wallis test). Pachymetry and Corneal Volume Table 5 shows the corneal pachymetry and volume data by group. There were statistically significant differences between the groups in the minimum and central pachymetry and in corneal volume (P%.04, 1-way ANOVA and Kruskal-Wallis tests). The central and minimum pachymetry values were statistically significantly lower in the keratoconus 2 and keratoconus 1 groups than in the subclinical and control groups (P%.03, Bonferroni test). There were no statistically significant differences in central pachymetry and minimum pachymetry between the subclinical
group and the keratoconus 1 group (P Z .99, Bonferroni test). There was a statistically significant difference in corneal volume between the control group and the keratoconus 2 group (P Z .01, Mann-Whitney test) and between the keratoconus 2 group and the keratoconus 1 group (P Z .03, Mann-Whitney test). Receiver Operating Characteristic The area under the ROC curve was statistically significant for all parameters (P!.01) except corneal volume (P Z .07). This area was higher than 0.5 for the anterior keratometric readings in the 3.0 mm zone (flattest meridian, 0.87; steepest meridian, 1.00; mean corneal power, 0.98) and for anterior astigmatism (0.94) and posterior astigmatism (0.92). The parameters showed the best sensitivity and specificity scores (sensitivity O76%; specificity O80%). In contrast, in the ROC analysis for detection of subclinical keratoconus, the anterior astigmatism and posterior astigmatism parameters provided an acceptable diagnostic performance, with an area under the ROC curve that was statistically significant and higher than 0.5 (Figure 7). The cutoff point was 1.05 D (sensitivity 85.7%; specificity 65%) for anterior astigmatism and 0.35 D (sensitivity 71%; specificity 60%) for posterior astigmatism.
J CATARACT REFRACT SURG - VOL 36, MAY 2010
CLINICAL AND SUBCLINICAL KERATOCONUS
821
Figure 5. Scattergrams of the relationship between the anterior mean keratometry and posterior mean keratometry. The adjusting line to the data obtained by means of the least-squares fit is shown in all graphs. A: Control group: Posterior mean keratometry Z 0.18 Anterior mean keratometry C1.46 (r2 Z 0.887). B: Subclinical group: Posterior mean keratometry Z 0.17 Anterior mean keratometry C1.24 (r2 Z 0.605). C: Keratoconus 1 group: Posterior mean keratometry Z 0.10 Anterior mean keratometry 2.36 (r2 Z 0.085). D: Keratoconus 2 group: Posterior mean keratometry Z 0.16 Anterior mean keratometry C0.46 (r2 Z 0.317).
DISCUSSION In the past few years, there has been much effort to use corneal topography technology to facilitate the diagnosis of subclinical keratoconus. One step toward that goal is the development of devices that evaluate the posterior corneal surface. The first commercially available topography system providing curvature and elevation data from the posterior corneal surface was based on a combination of scanning-slit and Placido-disk technologies. Several studies that used these technologies3–5,9–11 report changes in the posterior corneal surface in eyes with clinical or subclinical keratoconus. In 1 study,9 deformation, including local protrusion, occurred not only in the anterior corneal surface but also in the posterior corneal surface. Despite the evidence in these studies, concern remains
about the accuracy of posterior corneal surface measurements by scanning-slit and Placido-disk devices. Part of this controversy stems from problems with the reliability and reproducibility of some posterior corneal surface measurements in eyes that have had LASIK.12–16 For example, some posterior radially asymmetric measurements in these eyes, such as eccentricity, as well as measurements in eyes with corneal opacities, are not reliable with devices that use combined scanning-slit and Placido-disk technologies.12 The Scheimpflug photography–based system used in the present study provides repeatable pachymetric and posterior corneal surface (ie, curvature, elevation, asphericity) measurements.23–26 For this reason, we decided to use it to characterize posterior corneal
J CATARACT REFRACT SURG - VOL 36, MAY 2010
822
CLINICAL AND SUBCLINICAL KERATOCONUS
Figure 6. Scattergrams of the relationship between anterior mean astigmatism and posterior mean astigmatism. The adjusting line to the data obtained by means of the least-squares fit is shown in all graphs. A: Control group: Posterior mean astigmatism Z 0.18 Anterior mean astigmatism C0.13 (r2 Z 0.636). B: Subclinical group: Posterior mean astigmatism Z 0.23 Anterior mean astigmatism C0.11 (r2 Z 0.730). C: Keratoconus 1 group: Posterior mean astigmatism Z 0.19 Anterior mean astigmatism C0.21 (r2 Z 0.650). D: Keratoconus 2 group: Posterior mean astigmatism Z 0.19 Anterior mean astigmatism C0.14 (r2 Z 0.824).
changes in the current study of normal eyes (control group) and in eyes with subclinical keratoconus (subclinical group), keratoconus grade I (keratoconus 1 group), or keratoconus grade II (keratoconus 2 group). We also evaluated changes in the anterior corneal surface as well as the degree of correlation between the 2 corneal surfaces in the early to later stages of keratoconus and its contribution to pachymetry and corneal volume changes. Using the Scheimpflug system, we confirmed and characterized posterior surface changes and found a correlation between the changes and anterior surface, pachymetric, and corneal volume changes in eyes with keratoconus, even in subclinical cases. As expected, there were significant refractive differences between the 4 groups. Eyes with clinical keratoconus had the highest spherical equivalent and
cylinder values, whereas eyes with subclinical keratoconus had, on average, more significant manifest astigmatism than the control group. This higher cylindrical component in the subclinical group, although not as significant as in the 2 clinical keratoconus groups, has been reported.5 The anterior and posterior curvature changes were significantly in eyes with clinical keratoconus than in normal eyes. However, there were no significant differences in anterior or posterior central corneal curvature between the subclinical group and control group. This finding corroborates our definition of subclinical keratoconus; that is, no significant corneal steepening centrally but topographic asymmetry. Therefore, corneal topographic changes in eyes with subclinical keratoconus may begin peripherally (mainly in the inferior area), inducing corneal asymmetry. Tomidokoro et al.9 report
J CATARACT REFRACT SURG - VOL 36, MAY 2010
823
CLINICAL AND SUBCLINICAL KERATOCONUS
Table 5. Pachymetry and corneal volume by group. Group Parameter Pachymetry (mm) Central Mean G SD Range Minimum Mean G SD Range Corneal volume (mm2) Mean G SD Range
Control
Subclinical
Keratoconus 1
Keratoconus 2
549.90 G 28.48 483 to 598
514.29 G 43.59 448 to 613
501.63 G 32.76 429 to 555
457.61 G 38.77 402 to 516
547.55 G 28.71 482 to 595
507.00 G 46.83 426 to 610
492.32 G 36.77 418 to 548
446.44 G 34.58 399 to 504
60.83 G 3.27 54.90 to 66.70
58.91 G 4.87 52.30 to 69.50
60.00 G 2.84 55.30 to 64.20
57.98 G 2.65 52.80 to 60.60
P Value
!.01
!.01
.04
the same finding. Corneal astigmatism analysis in the central area showed significant differences between the control group and subclinical group as well as between the control group and the 2 clinical keratoconus groups. The higher anterior and posterior corneal astigmatism in the subclinical group than in the control group is consistent with the higher manifest cylinder values in the subclinical group. Schlegel et al.5 also found significantly increased anterior and posterior
Figure 7. The ROC curves for detection of subclinical keratoconus (discrimination between subclinical keratoconus and normal eyes) using the Scheimpflug system parameters showing an area under the curve higher than 0.5. The curves are separated from the diagonal; the nearer from the upper left corner of the graph, the larger the area under the ROC curve. A large area under the curve means that the false negative rate is high and the false positive rate is low and the sensitivity and specificity are acceptable.
corneal toricity in eyes with subclinical keratoconus compared with that in normal eyes. The only significant anterior and posterior BFS findings were lower values in more advanced cases of keratoconus (keratoconus 2 group) than in the control, subclinical, and keratoconus 1 groups. This parameter represents the total curvature of the cornea with a numeric value obtained by a least-squares adjustment; therefore, it is not sensitive enough to detect differences between different stages of keratoconus. In a study by de Sanctis et al.3 using the Scheimpflug imaging system, the mean maximum posterior corneal elevation in the 5.0 mm central area was significantly higher in eyes with clinical or subclinical keratoconus. However, to our knowledge, no study has reported the repeatability and reproducibility for maximum elevation. Thus, in the present study, we analyzed only the parameters with consistently confirmed repeatability. Regarding corneal asphericity, both corneal surfaces in eyes with keratoconus had a more significant prolate shape (negative asphericity). Significant differences were also found between the 2 clinical keratoconus groups; the most negative values were in eyes with more advanced keratoconus (grade II). We also evaluated the relationship between the geometric profile of the anterior and posterior corneal surfaces. We found a strong correlation between the anterior and posterior central curvatures in the control and subclinical groups; however, this correlation was lost in eyes with clinical keratoconus. Therefore, in cases of ectasia, the curvature of both corneal surfaces is affected but does not necessarily follow the same course. Using a combined scanning-slit–Placido system, Tomidokoro et al.9 found a lower correlation between anterior and posterior corneal curvatures in eyes with clinical or subclinical keratoconus than in normal eyes. It seems that the correlation between the anterior curvature and posterior curvature of the cornea is minimized in eyes with keratoconus. As
J CATARACT REFRACT SURG - VOL 36, MAY 2010
824
CLINICAL AND SUBCLINICAL KERATOCONUS
expected, analysis of the posterior–anterior corneal power ratio showed that the highest dioptric power was at the anterior corneal surface. Despite the poorer correlation between the anterior and posterior central curvatures in the keratoconus 2 group, that group had significantly higher posterior–anterior corneal power ratio values. This indicates that changes in posterior curvature were, on average, higher in magnitude in eyes with keratoconus grade II. Regarding corneal astigmatism, the correlation between the anterior surface and posterior surface was good in the subclinical, keratoconus 1, and keratoconus 2 groups. However, the correlation was moderate in the control group. This agrees with the findings of Dubbelman et al.,17 who found that both corneal surfaces tended to be flatter horizontally than vertically, and thus there is a degree of correlation in astigmatism between the 2 corneal surfaces. The analysis of the posterior–anterior astigmatism ratio showed the same tendency (ie, a more astigmatic anterior corneal surface) in all groups. Although the anterior and posterior BFS values were strongly correlated in the control and subclinical groups, the correlation was weaker in the 2 clinical keratoconus groups. This was consistent with the poor correlation in central curvature between the 2 surfaces of ectatic corneas. When a conic function was used for adjustment (asphericity), a moderate correlation between anterior corneal shape and posterior corneal shape was found in the subclinical and keratoconus 2 groups. Correlation was very good between the anterior corneal asphericity and posterior corneal asphericity in the keratoconus 1 group. However, the aspheric profile of both surfaces of normal corneas was not correlated, although central curvature was. Schlegel et al.5 also found poor correlations between anterior asphericity and posterior asphericity in normal eyes as well as in eyes with subclinical keratoconus. Finally, we evaluated the differences in pachymetry and corneal volume between groups. Pachymetry and analysis of corneal topographic patterns are considered indispensable tools in the preoperative screening of refractive surgery candidates and in the diagnosis of corneal ectasia.27 In our study, we found progressively lower pachymetric readings in subclinical, early, and moderate keratoconus cases, with the lowest values in the latter group. Therefore, corneal thinning in eyes with ectatic corneal disease can be accurately monitored using the Scheimpflug imaging system. Emre et al.28 obtained similar outcomes using the same system. Corneal volume as a new index to diagnose keratoconus and screen refractive candidates has also been proposed.29 However, we found no significant differences in corneal volume between the 4
groups. Eyes with keratoconus grade II had a significantly lower corneal volume that was consistent with the significant reduction in pachymetry. In eyes with keratoconus grade I, corneal thinning was observed, although a reduction in corneal volume was not. One likely explanation for this finding could be that in early stages of keratoconus, a redistribution of corneal volume occurs with no loss of tissue. In any case, this is something that should be addressed in future studies with a larger number of cases. We found that some parameters were useful in the detection of subclinical and clinical keratoconus. Specifically, the anterior keratometric and anterior and posterior corneal astigmatism values had acceptable sensitivity and specificity. Thus, anterior keratometry is crucial in the diagnosis of keratoconus. Adjustment parameters using a reference pattern (sphere, conic), such as asphericity and BFS, were poor predictors of keratoconus. A possible explanation is that adjusting corneal contour using a general curve leaves out information from small areas of corneal irregularity and this information is required for detecting the initial stages of keratoconus. The only useful factors in detecting subclinical keratoconus were anterior astigmatism and posterior astigmatism, although both had limited specificity. In summary, eyes with clinical or subclinical keratoconus had higher levels of manifest astigmatism and anterior and posterior corneal astigmatism. There was a correlation between the magnitude of astigmatism at the anterior and posterior corneal surfaces in all eyes, including normal eyes. In eyes with keratoconus, both corneal surfaces had a significant prolate shape in concordance with the steepening at both surfaces. On the other hand, keratometric changes at the anterior corneal surface were highly correlated with those at the posterior surface in normal and subclinical cases; the correlation was poorer in eyes with clinical keratoconus. These findings show that the correlation between the anterior curvature and the posterior curvature of the cornea is lost when keratoconus is present. In addition, changes in posterior curvature were, on average, higher in magnitude than those in the anterior surface in eyes with more advanced stages of keratoconus. All geometric changes at both corneal surfaces in keratoconus were accompanied by a reduction in pachymetry values and corneal volume, especially in more advanced cases. Therefore, the correlation between anterior curvature and posterior curvature no longer exists in cases of keratoconus and there is a correlation between the magnitude of astigmatism at the anterior and posterior corneal surfaces in normal eyes and in eyes with an ectatic cornea. Although these conclusions may seem contradictory, they are not. The astigmatism at the anterior surface can be high if the posterior astigmatism
J CATARACT REFRACT SURG - VOL 36, MAY 2010
825
CLINICAL AND SUBCLINICAL KERATOCONUS
is also high; however, the 2 surfaces may not have the same mean curvature. For example, a group of corneas with 3.00 D of astigmatism may not all have a mean keratometry of 49.00 D. For a more accurate diagnosis of subclinical and clinical keratoconus, postoperative corneal topography should be considered as well as anterior corneal topography. The cornea must be viewed as a 3-D element formed by 2 surfaces and having a specific volume. In the future, new diagnostic criteria should be developed using the 3-D characterization of the cornea because currently available devices use this model. For example, the potential of the relationship between anterior and posterior corneal surfaces as a new diagnostic tool should be evaluated with a larger keratoconus population to define a precise diagnostic criterion for this factor. REFERENCES 1. Rabinowitz YS. Keratoconus. Surv Ophthalmol 1998; 42:297–319 2. Wilson SE, Lin DTC, Klyce SD. Corneal topography of keratoconus. Cornea 1991; 10:2–8 3. de Sanctis U, Loiacono C, Richiardi L, Turco D, Mutani B, Grignolo FM. Sensitivity and specificity of posterior corneal elevation measured by Pentacam in discriminating keratoconus/ subclinical keratoconus. Ophthalmology 2008; 115:1534–1539 4. Nilforoushan MR, Speaker M, Marmor M, Abramson J, Tullo W, Morschauser D, Latkany R. Comparative evaluation of refractive surgery candidates with Placido topography, Orbscan II, Pentacam, and wavefront analysis. J Cataract Refract Surg 2008; 34:623–631 5. Schlegel Z, Hoang-Xuan T, Gatinel D. Comparison of and correlation between anterior and posterior corneal elevation maps in normal eyes and keratoconus-suspect eyes. J Cataract Refract Surg 2008; 34:789–795 6. Bu¨hren J, Ku¨hne C, Kohnen T. Defining subclinical keratoconus using corneal first-surface higher-order aberrations. Am J Ophthalmol 2007; 143:381–389 7. Jafri B, Li X, Yang H, Rabinowitz YS. Higher order aberrations and topography in early and suspected keratoconus. J Refract Surg 2007; 23:774–781 8. Lim L, Wei RH, Chan WK, Tan DTH. Evaluation of higher order ocular aberrations in patients with keratoconus. J Refract Surg 2007; 23:825–828 9. Tomidokoro A, Oshika T, Amano S, Higaki S, Maeda N, Miyata K. Changes in anterior and posterior corneal curvatures in keratoconus. Ophthalmology 2000; 107:1328–1332 10. Sonmez B, Doan M-P, Hamilton R. Identification of scanning slitbeam topographic parameters important in distinguishing normal from keratoconic corneal morphologic features. Am J Ophthalmol 2007; 143:401–408 11. Rao SN, Raviv T, Majmudar PA, Epstein RJ. Role of Orbscan II in screening keratoconus suspects before refractive corneal surgery. Ophthalmology 2002; 109:1642–1646 12. Maldonado MJ, Nieto JC, Dı´ez-Cuenca M, Pin˜ero DP. Repeatability and reproducibility of posterior corneal curvature measurements by combined scanning-slit and Placido-disc topography after LASIK. Ophthalmology 2006; 113:1918–1926 13. Ueda T, Nawa Y, Masuda K, Ishibashi H, Hara Y, Uozato H. Posterior corneal surface changes after hyperopic laser in situ keratomileusis. J Cataract Refract Surg 2005; 31:2084–2087
14. Nawa Y, Masuda K, Ueda T, Hara Y, Uozato H. Evaluation of apparent ectasia of the posterior surface of the cornea after keratorefractive surgery. J Cataract Refract Surg 2005; 31:571–573 15. Cairns G, Ormonde SE, Gray T, Hadden OB, Morris T, Ring P, McGhee CNJ. Assessing the accuracy of Orbscan II postLASIK: apparent keratectasia is paradoxically associated with anterior chamber depth reduction in successful procedures. Clin Exp Ophthalmol 2005; 33:147–152 16. Cairns G, McGhee CNJ. Orbscan computerized topography: attributes, applications, and limitations. J Cataract Refract Surg 2005; 31:205–220 17. Dubbelman M, Sicam VADP, Van der Heijde GL. The shape of the anterior and posterior surface of the aging human cornea. Vision Res 2006; 46:993–1001 18. Kawamorita T, Uozato H, Kamiya K, Bax L, Tsutsui K, Aizawa D, Shimizu K. Repeatability, reproducibility, and agreement characteristics of rotating Scheimpflug photography and scanningslit corneal topography for corneal power measurement. J Cataract Refract Surg 2009; 35:127–133 19. Alio´ JL, Shabayek MH. Corneal higher order aberrations: a method to grade keratoconus. J Refract Surg 2006; 22:539–545 20. Lim L, Wei RH, Chan WK, Tan DTH. Evaluation of keratoconus in Asians: role of Orbscan II and Tomey TMS-2 corneal topography. Am J Ophthalmol 2007; 143:390–400 21. Fan H-B, Lim K- L. Corneal elevation indices in normal and keratoconic eyes. J Cataract Refract Surg 2006; 32:1281–1287 22. Waring GO III. Nomenclature for keratoconus suspects [opinion]. Refract Corneal Surg 1993; 9:219–222 23. Shankar H, Taranath D, Santhirathelagan CT, Pesudovs K. Anterior segment biometry with the Pentacam: comprehensive assessment of repeatability of automated measurements. J Cataract Refract Surg 2008; 34:103–113 24. Pin˜ero DP, Saenz Gonza´lez C, Alio´ JL. Intraobserver and interobserver repeatability of curvatura and aberrometric measurements of the posterior corneal surface in normal eyes using Scheimpflug photography. J Cataract Refract Surg 2009; 35:113–120 25. Chen D, Lam AKC. Reliability and repeatability of the Pentacam on corneal curvatures. Clin Exp Optom 2009; 92:110–118 26. Chen D, Lam AKC. Intrasession and intersession repeatability of the Pentacam system on posterior corneal assessment in the normal human eye. J Cataract Refract Surg 2007; 33:448–454 27. Ambro´sio R Jr, Klyce SD, Wilson SE. Corneal topographic and pachymetric screening of keratorefractive patients. J Refract Surg 2003; 19:24–29 28. Emre S, Doganay S, Yologlu S. Evaluation of anterior segment parameters in keratoconic eyes measured with the Pentacam system. J Cataract Refract Surg 2007; 33:1708–1712 29. Ambrosio R Jr, Alonso RS, Luz A, Coca Velarde LG. Cornealthickness spatial profile and corneal-volume distribution: tomographic indices to detect keratoconus. J Cataract Refract Surg 2006; 32:1851–1859
J CATARACT REFRACT SURG - VOL 36, MAY 2010
First author David P. Pin˜ero, MSc Keratoconus Unit and Department of Optics, Vissum Corporation-Instituto Oftalmologico de Alicante, Alicante, Spain
Corneal volume, pachymetry, and correlation of anterior and posterior corneal shape in subclinical and different stages of clinical keratoconus David P. Pin˜ero, MSc, Jorge L. Alio´, MD, PhD, Alicia Aleso´n, OD, Munir Escaf Vergara, MD, Mauricio Miranda, MD
PURPOSE: To evaluate the corneal volume, pachymetry, and correlation of anterior and posterior corneal shape in subclinical and clinical keratoconus. SETTING: Vissum Corporation, Alicante, Spain. METHODS: Eyes were placed into 1 of 4 groups as follows: keratoconus 2 (grade II), keratoconus 1 (grade I), subclinical (subclinical keratoconus), and control (normal eyes). All eyes had an ophthalmologic examination including corneal evaluation (curvature, elevation, asphericity, pachymetry, corneal volume) by rotating Scheimpflug imaging (Pentacam). The posterior–anterior corneal power ratio was also calculated. RESULTS: Seventy-one eyes (51 patients; aged 16 to 64 years) were evaluated. Astigmatism and keratometry of both corneal surfaces were statistically significantly higher in the keratoconus 1 and 2 groups (P%.02). Posterior astigmatism was statistically significantly higher in the subclinical group than in the control group (P Z .01). A strong correlation (rR 0.81) was found between anterior and posterior curvature in the normal and subclinical groups; the correlation was weaker in clinical keratoconus cases (r% 0.56). The correlation in astigmatism between the anterior and posterior surface was good in all keratoconus groups (rR0.81). The posterior–anterior corneal power ratio was significantly higher in the keratoconus 2 group than in the other groups (P%.01). Pachymetric readings were progressively lower in eyes with subclinical, early, or moderate keratoconus (P<.01). The corneal volume was statistically significantly lower in the keratoconus 2 group than in the other groups (P Z .04). CONCLUSION: The correlation between anterior and posterior corneal curvature was lower in keratoconus, although the correlation between anterior and posterior astigmatism was maintained. Financial Disclosure: No author has a financial or proprietary interest in any material or method mentioned. J Cataract Refract Surg 2010; 36:814–825 Q 2010 ASCRS and ESCRS
Keratoconus is an ectatic corneal disorder characterized in most cases by progressive corneal thinning that results in corneal protrusion, irregular astigmatism, and decreased vision.1 Its incidence varies depending on several factors, such as ethnicity and the criteria used to establish the diagnosis; most estimates place the incidence in the general population between 50 and 230 per 100 000.1 Corneal topography is a valuable tool for confirming the keratoconus diagnosis.1 Significant corneal steepening in the anterior corneal surface of keratoconic eyes is always observed; the steepening is usually confined to 1 or 2 quadrants.1,2 Therefore, detecting moderate and advanced 814
Q 2010 ASCRS and ESCRS Published by Elsevier Inc.
keratoconus is not difficult using corneal topography and biomicroscopic, retinoscopic, and pachymetric findings.1 Detection is difficult with very early or preclinical stages of this ectatic disorder. The terms forme fruste keratoconus, subclinical keratoconus, and keratoconus suspect have been used to designate early stages of keratoconus that do not show on biomicroscopy but have subtle topographic features similar to those of clinical keratoconus.3–8 The biomicroscopic findings in keratoconus, such as stromal thinning and posterior stress lines, suggest that the posterior corneal shape is altered, perhaps independently of the anterior surface. Tomidokoro et al.9 0886-3350/10/$dsee front matter doi:10.1016/j.jcrs.2009.11.012
CLINICAL AND SUBCLINICAL KERATOCONUS
found that both anterior curvature and posterior curvature were affected in eyes with keratoconus as well as in keratoconus-suspect eyes. Significantly large best-fit sphere (BFS) values3,5,10 and high posterior elevations4,5,11 have been observed in eyes with clinical or subclinical keratoconus. Thus, manifestations of keratoconus occur at the posterior corneal surface, even in early stages of the disease, and identifying these changes could help clinicians detect subclinical keratoconus. To date, studies of posterior changes in eyes with keratoconus were performed using scanning-slit technology, which has been shown to be inaccurate for some posterior corneal measurements in eyes that have had laser in situ keratomileusis (LASIK).12–16 A fairly recent advance in corneal topography was the introduction of Scheimpflug photography for corneal topographic characterization; this technique allows the study of both the anterior and posterior corneal surfaces17 and provides more repeatable and reproducible anterior and posterior measurements of corneal power than scanning-slit technology. Kawamorita et al.18 reported 0.19 of agreement for withinrater consecutive measurements of posterior corneal power measurements of diopter (D) with a Scheimpflug-based system and 0.96 D with a scanning-slit system and for between-rater measurements, of 0.56 D and 1.58 D, respectively. The aim of the present study was to evaluate changes in the anterior and posterior corneal surfaces, pachymetry, and corneal volume in eyes with subclinical keratoconus or grade I or II clinical keratoconus using a Scheimpflug imaging system. We also analyzed the degree of correlation between the 2 corneal surfaces from the early stages of the disorder and
Submitted: March 4, 2009. Final revision submitted: November 20, 2009. Accepted: November 23, 2009.
815
the contribution of the correlation to volumetric modifications. PATIENTS AND METHODS This comparative study enrolled patients examined at Vissum Instituto Oftalmolo´gico de Alicante, Spain, or Instituto Oftalmosalud, Peru, between January 2008 and July 2008. All patients were informed about inclusion in the study and provided informed consent in accordance with the Declaration of Helsinki. An ethics board committee approved the study.
Patient Grouping Based on an examination, eyes were placed into 1 of 4 groups as follows: keratoconus 2 (grade II), keratoconus 1 (grade I), subclinical (subclinical keratoconus), and control (normal eyes). Clinical keratoconus was defined as evident findings characteristic of keratoconus (eg, corneal topography with asymmetric bow-tie pattern with or without skewed axes) (Figures 1 and 2), and at least 1 keratoconus sign (eg, stromal thinning, conical protrusion of the cornea at the apex, Fleischer ring, Vogt striae, or anterior stromal scar) on slitlamp examination.1 The Alio´ and Shabayek classification19 was used to grade the keratoconus. In the classification, grade II keratoconus (Figure 1) is characterized by a mean keratometry (K) value less than 53.00 D and a coma-like (ie, 3rd-, 5th-, and 7th-order Zernike terms) root-mean-square (RMS) value between 2.5 mm and 3.5 mm. Grade I keratoconus (Figure 2) is characterized by a mean K value less than 48.00 D and a coma-like RMS value between 1.50 mm and 2.50 mm. Subclinical keratoconus was diagnosed using criteria defined in previous studies,3–10,18,20–22 including corneal topography with abnormal localized steepening or an asymmetric bow-tie pattern (Figure 3), a normal-appearing cornea on slitlamp biomicroscopy, and at least 1 of the following signs: steep keratometric curvature (O47.00 D), oblique cylinder greater than 1.50 D, central corneal thickness less than 500 mm, or clinical keratoconus in the fellow eye. Eyes were considered normal if they had no ocular pathology, no previous ocular surgery, no significant refractive error, and no irregular corneal pattern (Figure 4). In this group, only 1 eye of each patient was evaluated (random sampling).
Patient Assessment
From the Keratoconus Unit (Pin˜ero, Alio´, Aleso´n), Vissum Corporation–Instituto Oftalmologico de Alicante, Departamento de O´ptica (Pin˜ero), Farmacologı´a y Anatomı´a, Universidad de Alicante, Division of Ophthalmology (Alio´), Universidad Miguel Herna´ndez, Alicante, Spain; Instituto Oftalmosalud SRL (Vergara, Miranda), Lima, Peru´. Supported in part by a grant from the Spanish Ministry of Health, Instituto Carlos III, Red Tema´tica de Investigacio´n Cooperativa en Salud Patologı´a Ocular del Envejecimiento, Calidad Visual y Calidad de Vida, Subproyecto de Calidad Visual (RD07/0062). Corresponding author: David P. Pin˜ero, MSc, Vissum Corporation– Instituto Oftalmologico de Alicante, Avenida de Denia s/n, Edificio Vissum, 03016 Alicante, Spain. E-mail: [email protected].
All patients had a complete ophthalmologic examination that included uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), manifest refraction, and slitlamp and fundus evaluations. Corneal evaluation was performed using a Scheimpflug imaging system (Pentacam, Oculus Optikgera¨te GmbH). This noninvasive system measures and characterizes the anterior segment.23 A rotating Scheimpflug camera captures 100 images with 500 measurement points on the anterior and posterior corneal surface over a 180-degree rotation. The elevation data from all images are combined to form a 3-dimensional (3-D) reconstruction of the corneal structure. After processing the information, the internal software provides several calculations and parameters. In the current study, software version 6.02r10 was used with a previously described measurement protocol.24 Several studies23–26 report that the system’s
J CATARACT REFRACT SURG - VOL 36, MAY 2010
816
CLINICAL AND SUBCLINICAL KERATOCONUS
Figure 1. Corneal topographic analysis of an eye with keratoconus grade II (Scheimpflug system). Left: Anterior axial topographic map. Right: Posterior axial topographic maps. The corneal geometry of the anterior corneal surface and posterior corneal surface is similar, with the same topographic pattern.
pachymetric and geometric measurements (curvature, asphericity, elevation) have good repeatability. The following anterior and posterior corneal surface parameters were evaluated with the Scheimpflug system: corneal dioptric power in the flattest meridian in the 3.0 mm central zone, corneal dioptric power in the steepest meridian in the 3.0 mm central zone, mean corneal power in the 3.0 mm zone, corneal astigmatism in the 3.0 mm zone, the BFS in an 8.0 mm diameter analysis area, and mean asphericity in an 8.0 mm diameter corneal area. Central pachymetry, minimum pachymetry, and corneal volume were also recorded and analyzed. In addition, the posterior–anterior corneal power ratio and posterior– anterior astigmatism ratio were calculated; that is, the ratios of the anterior value to the posterior value of the corresponding variable (ie, corneal power or astigmatism).
Statistical Analysis The number of patients in the study was chosen based on sample-size calculations obtained using a standard procedure. Considering means and standard deviations in previous studies3,5,9,17,23,24 of keratoconus diagnosis and assuming a specific significance level and statistical power (80%), the sample-size was estimated using Ene 2.0 software (GlaxoSmithKline). Statistical analysis was performed using SPSS software for Windows (version 15.0, SPSS, Inc.). Normality of all data
samples was first checked by the Shapiro-Wilks test. When parametric analysis was possible, 1-way analysis of variance (ANOVA) with Bonferroni post hoc comparison was used to compare visual, refractive, topographic, pachymetric, and volumetric parameters between the 4 groups of eyes. If variances were not homogeneous (checked by Levene test), Tamhane post hoc analysis was used. In all cases, differences were considered statistically significant when the P value was less than 0.05. When parametric analysis was not possible, the Kruskal-Wallis test was used for between-group comparisons and a P value less than 0.05 was considered statistically significant. Post hoc analysis was by the MannWhitney test with Bonferroni adjustment. Correlation coefficients (Pearson or Spearman depending on whether normality could be assumed) were used to assess the correlation between mean keratometry and astigmatism of the anterior corneal surface and posterior corneal surface in each of the 4 groups of eyes. In addition, linear regression analysis was performed to determine the relationship between anterior parameters and posterior parameters and to determine significant correlations between parameters. An analysis of the receiver operating characteristic (ROC) curve of Scheimpflug system parameters was performed to obtain cutoff values and thus detect keratoconus and keratoconus-suspect cases. This analysis was performed only for corneal parameters for which there was a statistically significant difference between control eyes and eyes with keratoconus or between control eyes and eyes with
Figure 2. Corneal topographic analysis of an eye with keratoconus grade I (Scheimpflug system). Left: Anterior axial topographic map. Right: Posterior axial topographic maps. The corneal geometry of the anterior corneal surface and posterior corneal surface is similar, with the same topographic pattern.
J CATARACT REFRACT SURG - VOL 36, MAY 2010
817
CLINICAL AND SUBCLINICAL KERATOCONUS
Figure 3. Corneal topographic analysis of an eye with subclinical keratoconus (Scheimpflug system). Left: Anterior axial topographic map. Right: Posterior axial topographic maps. The corneal geometry of the anterior corneal surface and posterior corneal surface is similar, with the same topographic pattern.
subclinical keratoconus. Cutoff values for detection of keratoconus (discrimination between eyes with keratoconus and control eyes) were calculated for all Scheimpflug system parameters except the ratios. Cutoff values to detect subclinical keratoconus cases (discrimination between eyes with subclinical keratoconus and normal eyes) were calculated for anterior and posterior corneal astigmatism in the 3.0 mm zone. These cutoff values corresponded to the points of the curve with the highest accuracy (ie, least false negative and false positive results). In addition, the area under the ROC curve was calculated to measure test accuracy, with an area of 1.0 representing a perfect test and an area of 0.5 representing an unusable test. The closer the curve follows the left-hand border and then the top border of the ROC space, the more accurate the test; that is, the test can identify more true positives while minimizing the number of false positives.
RESULTS The study evaluated 71 eyes of 51 patients ranging in age from 16 to 64 years. There were 37 right eyes (52.11%), 34 left eyes (47.89%), 29 men (56.86%), and 22 women (43.14%). No patient had a history of contact lens wear, eye disease, or surgery. The keratoconus 2 group comprised 18 eyes with a mean age of 39.06 years G 11.66 (SD). The
keratoconus 1 group comprised 19 eyes with a mean age of 33.58 G 11.26 years. The subclinical group comprised 14 eyes with a mean age of 30.00 G 9.15 years; all subclinical cases were unilateral. The control group comprised 20 eyes with a mean age of 32.30 G 6.59 years. No cone opacity was observed in any eye. Visual Acuity and Refraction Table 1 shows the visual acuity and refraction by group. Statistically significant differences in sphere and cylinder were found between groups (P%.01, Kruskal-Wallis test). Specifically, cylinder was statistically significantly higher in the keratoconus 1 and keratoconus 2 groups and sphere in the keratoconus 2 group (all P!.01, Mann-Whitney test). The only statistically significant difference between the control group and the subclinical group was in cylinder (P Z .03, Mann-Whitney test). There were no statistically significant differences in sphere or cylinder between the keratoconus 1 group and keratoconus 2 group (all PR.18, Mann-Whitney test). The difference in CDVA between the 4 groups was statistically significant (P!.01, Kruskal-Wallis test).
Figure 4. Corneal topographic analysis of a normal eye (Scheimpflug system). Left: Anterior axial topographic map. Right: Posterior axial topographic maps. There are differences between the anterior and posterior topographic patterns, especially in the magnitude of astigmatism.
J CATARACT REFRACT SURG - VOL 36, MAY 2010
818
CLINICAL AND SUBCLINICAL KERATOCONUS
Table 1. Visual acuity and refractive outcomes. Group Parameter Sphere (D) Mean G SD Range Cylinder (D) Mean G SD Range SE (D) Mean G SD Range CDVA Mean G SD Range
Control
Subclinical
Keratoconus 1
Keratoconus 2
1.26 G 2.54 6.00 to C4.00
1.16 G 1.15 3.25 to 0.00
3.13 G 4.19 C1.75 to 13.75
5.36 G 4.56 17.00 to C2.00
0.54 G 1.00 3.50 to 0.00
1.21 G 1.05 3.25 to 0.00
3.80 G 2.16 8.50 to 1.00
4.28 G 2.20 9.00 to 0.00
1.77 G 1.35 4.88 to 0.00
5.03 G 4.13 14.75 to C0.63
7.50 G 4.14 18.00 to 0.25
1.02 G 0.13 0.80 to 1.20
0.80 G 0.19 0.33 to 1.00
0.44 G 0.20 0.13 to 0.80
P Value .01
!.01
!.01 1.48 G 2.43 7.00 to C2.25
!.01 1.12 G 0.10 1.00 to 1.20
CDVA Z corrected distance visual acuity; SE Z spherical equivalent
Curvature Table 2 shows the curvature and elevation data for the anterior corneal surface by group. There were statistically significant between-group differences in the flattest and steepest meridians, mean corneal power, and corneal astigmatism in the 3.0 mm central zone; the BFS; and the mean asphericity in an 8.0 mm diameter area (P!.01, 1-way ANOVA and Kruskal-Wallis tests). Specifically, the corneal dioptric power in the steepest meridian and mean corneal power in the 3.0 mm central zone were statistically significantly higher in the keratoconus 2 and keratoconus 1 groups than in the subclinical and control groups (P!.01, Mann-Whitney test). The corneal dioptric power in the flattest meridian in the 3.0 mm central zone was significantly higher in the keratoconus 2 and keratoconus 1 groups than in the control and subclinical groups and in the keratoconus 1 group than in the control group (P!.01 and P Z .04, respectively; Bonferroni test). The anterior corneal astigmatism was statistically significantly higher in the keratoconus 2, keratoconus 1, and subclinical groups than in the control group (P!.01, Mann-Whitney test). The BFS value was statistically significant lower in the keratoconus 2 group (P!.01, Bonferroni test); the difference between the subclinical group and keratoconus 1 group was not significant (P Z .99, Bonferroni test). The anterior corneal surface was significantly more prolate in the keratoconus 2 group and keratoconus 1 group than in the other 2 groups (P!.01, Bonferroni test). Table 3 shows the curvature and elevation data for the posterior corneal surface by group. The corneal dioptric power in the flattest and steepest meridians, mean corneal power, and corneal astigmatism in the 3.0 mm central zone were statistically significantly
higher in the keratoconus 2 and keratoconus 1 groups than in the subclinical and control groups (P%.02, Bonferroni and Mann-Whitney tests). Statistically significant differences in posterior astigmatism between the control group and subclinical group were also found (P Z .01, Mann-Whitney test). The posterior BFS value was statistically significantly lower in the keratoconus 2 group than in the other 3 groups (P%.02, Mann-Whitney test); there were no significant differences in the BFS between the subclinical group and the keratoconus 1 group (P Z .24, MannWhitney test). The posterior corneal surface was significantly more prolate (marked negative asphericity) in the keratoconus 2 and keratoconus 1 groups than in the other 2 groups (P!.01, Mann-Whitney test). Relationship Between Anterior and Posterior Corneal Surfaces Table 4 shows the correlation coefficients between the curvatures, BFS, and asphericities of anterior and posterior corneal surfaces by group. In the control group, a significant and strong correlation was found between the anterior curvature and posterior curvature (Figure 5). However, the correlation was lower in the other 3 groups, especially in the keratoconus 1 group, in which the correlation between the anterior and posterior mean keratometric values was poor. Regarding astigmatism, there was a good correlation between the anterior surface and posterior surface in all groups except the control group (Figure 6). There was a strong correlation between the anterior BFS and posterior BFS in the control group and subclinical group. Regarding corneal asphericity, the correlation between the anterior surface and posterior surface was good in the keratoconus 1 group, moderate in
J CATARACT REFRACT SURG - VOL 36, MAY 2010
819
CLINICAL AND SUBCLINICAL KERATOCONUS
Table 2. Curvature, asphericity, and elevation data for the anterior corneal surface by group. Group Parameter Corneal power* (D) Flattest meridian Mean G SD Range Steepest meridian Mean G SD Range Mean corneal power* (D) Mean G SD Range Corneal astigmatism* (D) Mean G SD Range BFS† (mm) Mean G SD Range Mean asphericity† Mean G SD Range
Control
Subclinical
Keratoconus 1
Keratoconus 2
42.82 G 1.69 40.10 to 46.30
43.00 G 1.31 40.90 to 45.40
44.27 G 1.30 42.50 to 46.30
48.27 G 2.06 44.90 to 52.50
43.81 G 1.97 40.80 to 47.00
44.92 G 1.77 41.60 to 47.60
49.05 G 1.07 47.80 to 52.00
53.09 G 2.20 48.60 to 56.30
43.29 G 1.74 40.40 to 46.60
43.93 G 1.47 41.20 to 46.50
46.51 G 0.70 45.30 to 47.60
50.52 G 1.60 48.10 to 53.60
1.13 G 0.84 0.40 to 3.70
1.93 G 1.02 0.40 to 3.80
4.55 G 2.11 0.90 to 9.30
4.83 G 2.83 0.80 to 10.30
7.89 G 0.31 7.32 to 8.43
7.74 G 0.30 7.08 to 8.18
7.64 G 0.23 7.27 to 8.04
7.26 G 0.30 6.70 to 7.75
0.29 G 0.09 0.46 to 0.14
0.33 G 0.23 0.67 to 0.01
0.65 G 0.27 1.33 to 0.26
1.18 G 0.32 1.68 to 0.58
P Value
!.01
!.01
!.01
!.01
!.01
!.01
BFS Z best-fit sphere *Central 3.0 mm zone † 8.0 mm diameter area
Table 3. Curvature, asphericity and elevation data for the posterior corneal surface by group. Group Parameter Corneal power* (D) Flattest meridian Mean G SD Range Steepest meridian Mean G SD Range Mean corneal power* (D) Mean G SD Range Corneal astigmatism* (D) Mean G SD Range BFS† (mm) Mean G SD Range Mean asphericity† Mean G SD Range
Control
Subclinical
Keratoconus 1
Keratoconus 2
P Value
!.01 6.15 G 0.31 6.70 to 5.60
6.16 G 0.30 6.90 to 5.80
6.53 G 0.41 7.70 to 6.00
7.22 G 0.61 8.30 to 6.00
6.48 G 0.42 7.20 to 5.90
6.73 G 0.44 7.40 to 6.10
7.27 G 0.62 8.30 to 5.60
8.29 G 0.43 9.10 to 7.40
6.31 G 0.33 6.90 to 5.80
6.42 G 0.33 7.10 to 6.00
6.86 G 0.23 7.40 to 6.40
7.69 G 0.46 8.70 to 7.00
0.33 G 0.19 0.10 to 0.70
0.56 G 0.28 0.30 to 1.20
1.07 G 0.50 0.30 to 1.90
1.07 G 0.60 0.30 to 2.50
6.48 G 0.31 5.89 to 7.04
6.34 G 0.28 5.64 to 6.65
6.25 G 0.31 5.91 to 7.26
6.14 G 0.76 5.42 to 8.54
0.66 G 0.33 1.40 to 0.04
1.17 G 0.42 1.81 to 0.14
!.01
!.01
!.01
.08
!.01 0.34 G 0.24 1.01 to 0.02
0.36 G 0.33 1.03 to 0.05
BFS Z best-fit sphere *Central 3.0 mm zone † 8.0 mm diameter area
J CATARACT REFRACT SURG - VOL 36, MAY 2010
820
CLINICAL AND SUBCLINICAL KERATOCONUS
Table 4. Correlation coefficients for curvature, asphericity and elevation parameters between the anterior corneal surface and posterior corneal surface by group. Group Parameter Corneal power* (D) Flattest meridian r value P value Steepest meridian r value P value Mean corneal power* (D) r value P value Corneal astigmatism* (D) r value P value BFS† (mm) r value P value Mean asphericity† r value P value
Control
Subclinical
Keratoconus 1
Keratoconus 2
0.90 !.01
0.83 !.01
0.37 .12
0.73 .01
0.96 !.01
0.81 !.01
0.32 .19
0.67 !.01
0.94 !.01
0.85 !.01
0.34 .16
0.56 .02
0.62 !.01
0.85 !.01
0.81 .01
0.91 !.01
0.92 !.01
0.96 !.01
0.65 !.01
0.56 .02
0.17 .47
0.62 .02
0.89 !.01
0.63 !.01
BFS Z best-fit sphere *Central 3.0 mm zone † 8.0 mm diameter area
the subclinical and keratoconus 2 groups, and very poor in the control group (Table 4). The mean posterior–anterior corneal power ratio was 0.146 G 0.003 in the control group, 0.146 G 0.005 in the subclinical group, 0.148 G 0.005 in the keratoconus 1 group, and 0.152 G 0.007 in the keratoconus 2 group. The ratio was statistically significantly higher in the keratoconus 2 group than in the other 3 groups. The mean posterior–anterior astigmatism ratio was 0.333 G 0.170, 0.338 G 0.202, 0.268 G 0.188, and 0.257 G 0.134, respectively. There were no statistically significant differences in the ratio between the groups (P Z .12, Kruskal-Wallis test). Pachymetry and Corneal Volume Table 5 shows the corneal pachymetry and volume data by group. There were statistically significant differences between the groups in the minimum and central pachymetry and in corneal volume (P%.04, 1-way ANOVA and Kruskal-Wallis tests). The central and minimum pachymetry values were statistically significantly lower in the keratoconus 2 and keratoconus 1 groups than in the subclinical and control groups (P%.03, Bonferroni test). There were no statistically significant differences in central pachymetry and minimum pachymetry between the subclinical
group and the keratoconus 1 group (P Z .99, Bonferroni test). There was a statistically significant difference in corneal volume between the control group and the keratoconus 2 group (P Z .01, Mann-Whitney test) and between the keratoconus 2 group and the keratoconus 1 group (P Z .03, Mann-Whitney test). Receiver Operating Characteristic The area under the ROC curve was statistically significant for all parameters (P!.01) except corneal volume (P Z .07). This area was higher than 0.5 for the anterior keratometric readings in the 3.0 mm zone (flattest meridian, 0.87; steepest meridian, 1.00; mean corneal power, 0.98) and for anterior astigmatism (0.94) and posterior astigmatism (0.92). The parameters showed the best sensitivity and specificity scores (sensitivity O76%; specificity O80%). In contrast, in the ROC analysis for detection of subclinical keratoconus, the anterior astigmatism and posterior astigmatism parameters provided an acceptable diagnostic performance, with an area under the ROC curve that was statistically significant and higher than 0.5 (Figure 7). The cutoff point was 1.05 D (sensitivity 85.7%; specificity 65%) for anterior astigmatism and 0.35 D (sensitivity 71%; specificity 60%) for posterior astigmatism.
J CATARACT REFRACT SURG - VOL 36, MAY 2010
CLINICAL AND SUBCLINICAL KERATOCONUS
821
Figure 5. Scattergrams of the relationship between the anterior mean keratometry and posterior mean keratometry. The adjusting line to the data obtained by means of the least-squares fit is shown in all graphs. A: Control group: Posterior mean keratometry Z 0.18 Anterior mean keratometry C1.46 (r2 Z 0.887). B: Subclinical group: Posterior mean keratometry Z 0.17 Anterior mean keratometry C1.24 (r2 Z 0.605). C: Keratoconus 1 group: Posterior mean keratometry Z 0.10 Anterior mean keratometry 2.36 (r2 Z 0.085). D: Keratoconus 2 group: Posterior mean keratometry Z 0.16 Anterior mean keratometry C0.46 (r2 Z 0.317).
DISCUSSION In the past few years, there has been much effort to use corneal topography technology to facilitate the diagnosis of subclinical keratoconus. One step toward that goal is the development of devices that evaluate the posterior corneal surface. The first commercially available topography system providing curvature and elevation data from the posterior corneal surface was based on a combination of scanning-slit and Placido-disk technologies. Several studies that used these technologies3–5,9–11 report changes in the posterior corneal surface in eyes with clinical or subclinical keratoconus. In 1 study,9 deformation, including local protrusion, occurred not only in the anterior corneal surface but also in the posterior corneal surface. Despite the evidence in these studies, concern remains
about the accuracy of posterior corneal surface measurements by scanning-slit and Placido-disk devices. Part of this controversy stems from problems with the reliability and reproducibility of some posterior corneal surface measurements in eyes that have had LASIK.12–16 For example, some posterior radially asymmetric measurements in these eyes, such as eccentricity, as well as measurements in eyes with corneal opacities, are not reliable with devices that use combined scanning-slit and Placido-disk technologies.12 The Scheimpflug photography–based system used in the present study provides repeatable pachymetric and posterior corneal surface (ie, curvature, elevation, asphericity) measurements.23–26 For this reason, we decided to use it to characterize posterior corneal
J CATARACT REFRACT SURG - VOL 36, MAY 2010
822
CLINICAL AND SUBCLINICAL KERATOCONUS
Figure 6. Scattergrams of the relationship between anterior mean astigmatism and posterior mean astigmatism. The adjusting line to the data obtained by means of the least-squares fit is shown in all graphs. A: Control group: Posterior mean astigmatism Z 0.18 Anterior mean astigmatism C0.13 (r2 Z 0.636). B: Subclinical group: Posterior mean astigmatism Z 0.23 Anterior mean astigmatism C0.11 (r2 Z 0.730). C: Keratoconus 1 group: Posterior mean astigmatism Z 0.19 Anterior mean astigmatism C0.21 (r2 Z 0.650). D: Keratoconus 2 group: Posterior mean astigmatism Z 0.19 Anterior mean astigmatism C0.14 (r2 Z 0.824).
changes in the current study of normal eyes (control group) and in eyes with subclinical keratoconus (subclinical group), keratoconus grade I (keratoconus 1 group), or keratoconus grade II (keratoconus 2 group). We also evaluated changes in the anterior corneal surface as well as the degree of correlation between the 2 corneal surfaces in the early to later stages of keratoconus and its contribution to pachymetry and corneal volume changes. Using the Scheimpflug system, we confirmed and characterized posterior surface changes and found a correlation between the changes and anterior surface, pachymetric, and corneal volume changes in eyes with keratoconus, even in subclinical cases. As expected, there were significant refractive differences between the 4 groups. Eyes with clinical keratoconus had the highest spherical equivalent and
cylinder values, whereas eyes with subclinical keratoconus had, on average, more significant manifest astigmatism than the control group. This higher cylindrical component in the subclinical group, although not as significant as in the 2 clinical keratoconus groups, has been reported.5 The anterior and posterior curvature changes were significantly in eyes with clinical keratoconus than in normal eyes. However, there were no significant differences in anterior or posterior central corneal curvature between the subclinical group and control group. This finding corroborates our definition of subclinical keratoconus; that is, no significant corneal steepening centrally but topographic asymmetry. Therefore, corneal topographic changes in eyes with subclinical keratoconus may begin peripherally (mainly in the inferior area), inducing corneal asymmetry. Tomidokoro et al.9 report
J CATARACT REFRACT SURG - VOL 36, MAY 2010
823
CLINICAL AND SUBCLINICAL KERATOCONUS
Table 5. Pachymetry and corneal volume by group. Group Parameter Pachymetry (mm) Central Mean G SD Range Minimum Mean G SD Range Corneal volume (mm2) Mean G SD Range
Control
Subclinical
Keratoconus 1
Keratoconus 2
549.90 G 28.48 483 to 598
514.29 G 43.59 448 to 613
501.63 G 32.76 429 to 555
457.61 G 38.77 402 to 516
547.55 G 28.71 482 to 595
507.00 G 46.83 426 to 610
492.32 G 36.77 418 to 548
446.44 G 34.58 399 to 504
60.83 G 3.27 54.90 to 66.70
58.91 G 4.87 52.30 to 69.50
60.00 G 2.84 55.30 to 64.20
57.98 G 2.65 52.80 to 60.60
P Value
!.01
!.01
.04
the same finding. Corneal astigmatism analysis in the central area showed significant differences between the control group and subclinical group as well as between the control group and the 2 clinical keratoconus groups. The higher anterior and posterior corneal astigmatism in the subclinical group than in the control group is consistent with the higher manifest cylinder values in the subclinical group. Schlegel et al.5 also found significantly increased anterior and posterior
Figure 7. The ROC curves for detection of subclinical keratoconus (discrimination between subclinical keratoconus and normal eyes) using the Scheimpflug system parameters showing an area under the curve higher than 0.5. The curves are separated from the diagonal; the nearer from the upper left corner of the graph, the larger the area under the ROC curve. A large area under the curve means that the false negative rate is high and the false positive rate is low and the sensitivity and specificity are acceptable.
corneal toricity in eyes with subclinical keratoconus compared with that in normal eyes. The only significant anterior and posterior BFS findings were lower values in more advanced cases of keratoconus (keratoconus 2 group) than in the control, subclinical, and keratoconus 1 groups. This parameter represents the total curvature of the cornea with a numeric value obtained by a least-squares adjustment; therefore, it is not sensitive enough to detect differences between different stages of keratoconus. In a study by de Sanctis et al.3 using the Scheimpflug imaging system, the mean maximum posterior corneal elevation in the 5.0 mm central area was significantly higher in eyes with clinical or subclinical keratoconus. However, to our knowledge, no study has reported the repeatability and reproducibility for maximum elevation. Thus, in the present study, we analyzed only the parameters with consistently confirmed repeatability. Regarding corneal asphericity, both corneal surfaces in eyes with keratoconus had a more significant prolate shape (negative asphericity). Significant differences were also found between the 2 clinical keratoconus groups; the most negative values were in eyes with more advanced keratoconus (grade II). We also evaluated the relationship between the geometric profile of the anterior and posterior corneal surfaces. We found a strong correlation between the anterior and posterior central curvatures in the control and subclinical groups; however, this correlation was lost in eyes with clinical keratoconus. Therefore, in cases of ectasia, the curvature of both corneal surfaces is affected but does not necessarily follow the same course. Using a combined scanning-slit–Placido system, Tomidokoro et al.9 found a lower correlation between anterior and posterior corneal curvatures in eyes with clinical or subclinical keratoconus than in normal eyes. It seems that the correlation between the anterior curvature and posterior curvature of the cornea is minimized in eyes with keratoconus. As
J CATARACT REFRACT SURG - VOL 36, MAY 2010
824
CLINICAL AND SUBCLINICAL KERATOCONUS
expected, analysis of the posterior–anterior corneal power ratio showed that the highest dioptric power was at the anterior corneal surface. Despite the poorer correlation between the anterior and posterior central curvatures in the keratoconus 2 group, that group had significantly higher posterior–anterior corneal power ratio values. This indicates that changes in posterior curvature were, on average, higher in magnitude in eyes with keratoconus grade II. Regarding corneal astigmatism, the correlation between the anterior surface and posterior surface was good in the subclinical, keratoconus 1, and keratoconus 2 groups. However, the correlation was moderate in the control group. This agrees with the findings of Dubbelman et al.,17 who found that both corneal surfaces tended to be flatter horizontally than vertically, and thus there is a degree of correlation in astigmatism between the 2 corneal surfaces. The analysis of the posterior–anterior astigmatism ratio showed the same tendency (ie, a more astigmatic anterior corneal surface) in all groups. Although the anterior and posterior BFS values were strongly correlated in the control and subclinical groups, the correlation was weaker in the 2 clinical keratoconus groups. This was consistent with the poor correlation in central curvature between the 2 surfaces of ectatic corneas. When a conic function was used for adjustment (asphericity), a moderate correlation between anterior corneal shape and posterior corneal shape was found in the subclinical and keratoconus 2 groups. Correlation was very good between the anterior corneal asphericity and posterior corneal asphericity in the keratoconus 1 group. However, the aspheric profile of both surfaces of normal corneas was not correlated, although central curvature was. Schlegel et al.5 also found poor correlations between anterior asphericity and posterior asphericity in normal eyes as well as in eyes with subclinical keratoconus. Finally, we evaluated the differences in pachymetry and corneal volume between groups. Pachymetry and analysis of corneal topographic patterns are considered indispensable tools in the preoperative screening of refractive surgery candidates and in the diagnosis of corneal ectasia.27 In our study, we found progressively lower pachymetric readings in subclinical, early, and moderate keratoconus cases, with the lowest values in the latter group. Therefore, corneal thinning in eyes with ectatic corneal disease can be accurately monitored using the Scheimpflug imaging system. Emre et al.28 obtained similar outcomes using the same system. Corneal volume as a new index to diagnose keratoconus and screen refractive candidates has also been proposed.29 However, we found no significant differences in corneal volume between the 4
groups. Eyes with keratoconus grade II had a significantly lower corneal volume that was consistent with the significant reduction in pachymetry. In eyes with keratoconus grade I, corneal thinning was observed, although a reduction in corneal volume was not. One likely explanation for this finding could be that in early stages of keratoconus, a redistribution of corneal volume occurs with no loss of tissue. In any case, this is something that should be addressed in future studies with a larger number of cases. We found that some parameters were useful in the detection of subclinical and clinical keratoconus. Specifically, the anterior keratometric and anterior and posterior corneal astigmatism values had acceptable sensitivity and specificity. Thus, anterior keratometry is crucial in the diagnosis of keratoconus. Adjustment parameters using a reference pattern (sphere, conic), such as asphericity and BFS, were poor predictors of keratoconus. A possible explanation is that adjusting corneal contour using a general curve leaves out information from small areas of corneal irregularity and this information is required for detecting the initial stages of keratoconus. The only useful factors in detecting subclinical keratoconus were anterior astigmatism and posterior astigmatism, although both had limited specificity. In summary, eyes with clinical or subclinical keratoconus had higher levels of manifest astigmatism and anterior and posterior corneal astigmatism. There was a correlation between the magnitude of astigmatism at the anterior and posterior corneal surfaces in all eyes, including normal eyes. In eyes with keratoconus, both corneal surfaces had a significant prolate shape in concordance with the steepening at both surfaces. On the other hand, keratometric changes at the anterior corneal surface were highly correlated with those at the posterior surface in normal and subclinical cases; the correlation was poorer in eyes with clinical keratoconus. These findings show that the correlation between the anterior curvature and the posterior curvature of the cornea is lost when keratoconus is present. In addition, changes in posterior curvature were, on average, higher in magnitude than those in the anterior surface in eyes with more advanced stages of keratoconus. All geometric changes at both corneal surfaces in keratoconus were accompanied by a reduction in pachymetry values and corneal volume, especially in more advanced cases. Therefore, the correlation between anterior curvature and posterior curvature no longer exists in cases of keratoconus and there is a correlation between the magnitude of astigmatism at the anterior and posterior corneal surfaces in normal eyes and in eyes with an ectatic cornea. Although these conclusions may seem contradictory, they are not. The astigmatism at the anterior surface can be high if the posterior astigmatism
J CATARACT REFRACT SURG - VOL 36, MAY 2010
825
CLINICAL AND SUBCLINICAL KERATOCONUS
is also high; however, the 2 surfaces may not have the same mean curvature. For example, a group of corneas with 3.00 D of astigmatism may not all have a mean keratometry of 49.00 D. For a more accurate diagnosis of subclinical and clinical keratoconus, postoperative corneal topography should be considered as well as anterior corneal topography. The cornea must be viewed as a 3-D element formed by 2 surfaces and having a specific volume. In the future, new diagnostic criteria should be developed using the 3-D characterization of the cornea because currently available devices use this model. For example, the potential of the relationship between anterior and posterior corneal surfaces as a new diagnostic tool should be evaluated with a larger keratoconus population to define a precise diagnostic criterion for this factor. REFERENCES 1. Rabinowitz YS. Keratoconus. Surv Ophthalmol 1998; 42:297–319 2. Wilson SE, Lin DTC, Klyce SD. Corneal topography of keratoconus. Cornea 1991; 10:2–8 3. de Sanctis U, Loiacono C, Richiardi L, Turco D, Mutani B, Grignolo FM. Sensitivity and specificity of posterior corneal elevation measured by Pentacam in discriminating keratoconus/ subclinical keratoconus. Ophthalmology 2008; 115:1534–1539 4. Nilforoushan MR, Speaker M, Marmor M, Abramson J, Tullo W, Morschauser D, Latkany R. Comparative evaluation of refractive surgery candidates with Placido topography, Orbscan II, Pentacam, and wavefront analysis. J Cataract Refract Surg 2008; 34:623–631 5. Schlegel Z, Hoang-Xuan T, Gatinel D. Comparison of and correlation between anterior and posterior corneal elevation maps in normal eyes and keratoconus-suspect eyes. J Cataract Refract Surg 2008; 34:789–795 6. Bu¨hren J, Ku¨hne C, Kohnen T. Defining subclinical keratoconus using corneal first-surface higher-order aberrations. Am J Ophthalmol 2007; 143:381–389 7. Jafri B, Li X, Yang H, Rabinowitz YS. Higher order aberrations and topography in early and suspected keratoconus. J Refract Surg 2007; 23:774–781 8. Lim L, Wei RH, Chan WK, Tan DTH. Evaluation of higher order ocular aberrations in patients with keratoconus. J Refract Surg 2007; 23:825–828 9. Tomidokoro A, Oshika T, Amano S, Higaki S, Maeda N, Miyata K. Changes in anterior and posterior corneal curvatures in keratoconus. Ophthalmology 2000; 107:1328–1332 10. Sonmez B, Doan M-P, Hamilton R. Identification of scanning slitbeam topographic parameters important in distinguishing normal from keratoconic corneal morphologic features. Am J Ophthalmol 2007; 143:401–408 11. Rao SN, Raviv T, Majmudar PA, Epstein RJ. Role of Orbscan II in screening keratoconus suspects before refractive corneal surgery. Ophthalmology 2002; 109:1642–1646 12. Maldonado MJ, Nieto JC, Dı´ez-Cuenca M, Pin˜ero DP. Repeatability and reproducibility of posterior corneal curvature measurements by combined scanning-slit and Placido-disc topography after LASIK. Ophthalmology 2006; 113:1918–1926 13. Ueda T, Nawa Y, Masuda K, Ishibashi H, Hara Y, Uozato H. Posterior corneal surface changes after hyperopic laser in situ keratomileusis. J Cataract Refract Surg 2005; 31:2084–2087
14. Nawa Y, Masuda K, Ueda T, Hara Y, Uozato H. Evaluation of apparent ectasia of the posterior surface of the cornea after keratorefractive surgery. J Cataract Refract Surg 2005; 31:571–573 15. Cairns G, Ormonde SE, Gray T, Hadden OB, Morris T, Ring P, McGhee CNJ. Assessing the accuracy of Orbscan II postLASIK: apparent keratectasia is paradoxically associated with anterior chamber depth reduction in successful procedures. Clin Exp Ophthalmol 2005; 33:147–152 16. Cairns G, McGhee CNJ. Orbscan computerized topography: attributes, applications, and limitations. J Cataract Refract Surg 2005; 31:205–220 17. Dubbelman M, Sicam VADP, Van der Heijde GL. The shape of the anterior and posterior surface of the aging human cornea. Vision Res 2006; 46:993–1001 18. Kawamorita T, Uozato H, Kamiya K, Bax L, Tsutsui K, Aizawa D, Shimizu K. Repeatability, reproducibility, and agreement characteristics of rotating Scheimpflug photography and scanningslit corneal topography for corneal power measurement. J Cataract Refract Surg 2009; 35:127–133 19. Alio´ JL, Shabayek MH. Corneal higher order aberrations: a method to grade keratoconus. J Refract Surg 2006; 22:539–545 20. Lim L, Wei RH, Chan WK, Tan DTH. Evaluation of keratoconus in Asians: role of Orbscan II and Tomey TMS-2 corneal topography. Am J Ophthalmol 2007; 143:390–400 21. Fan H-B, Lim K- L. Corneal elevation indices in normal and keratoconic eyes. J Cataract Refract Surg 2006; 32:1281–1287 22. Waring GO III. Nomenclature for keratoconus suspects [opinion]. Refract Corneal Surg 1993; 9:219–222 23. Shankar H, Taranath D, Santhirathelagan CT, Pesudovs K. Anterior segment biometry with the Pentacam: comprehensive assessment of repeatability of automated measurements. J Cataract Refract Surg 2008; 34:103–113 24. Pin˜ero DP, Saenz Gonza´lez C, Alio´ JL. Intraobserver and interobserver repeatability of curvatura and aberrometric measurements of the posterior corneal surface in normal eyes using Scheimpflug photography. J Cataract Refract Surg 2009; 35:113–120 25. Chen D, Lam AKC. Reliability and repeatability of the Pentacam on corneal curvatures. Clin Exp Optom 2009; 92:110–118 26. Chen D, Lam AKC. Intrasession and intersession repeatability of the Pentacam system on posterior corneal assessment in the normal human eye. J Cataract Refract Surg 2007; 33:448–454 27. Ambro´sio R Jr, Klyce SD, Wilson SE. Corneal topographic and pachymetric screening of keratorefractive patients. J Refract Surg 2003; 19:24–29 28. Emre S, Doganay S, Yologlu S. Evaluation of anterior segment parameters in keratoconic eyes measured with the Pentacam system. J Cataract Refract Surg 2007; 33:1708–1712 29. Ambrosio R Jr, Alonso RS, Luz A, Coca Velarde LG. Cornealthickness spatial profile and corneal-volume distribution: tomographic indices to detect keratoconus. J Cataract Refract Surg 2006; 32:1851–1859
J CATARACT REFRACT SURG - VOL 36, MAY 2010
First author David P. Pin˜ero, MSc Keratoconus Unit and Department of Optics, Vissum Corporation-Instituto Oftalmologico de Alicante, Alicante, Spain